PDA

View Full Version : Enabling glslang in catalyst?



Myagi
09-21-2003, 02:00 AM
While searching this forum and the net I've seen that glslang functionality (even if it was experimental) was exposed in the ATI catalyst drivers, however it doesn't seem to be enabled in the latest driver.

I can see the functions when looking in the dll, they have been renamned from GL2 to ARB but other than that they're there. wglGetProcAddress returns NULL though.

I can't find any info at all on the net that relates to the latest 3.7 driver and how to get it to work. Does anyone know what registry key or whatever I have to do to "enable" it?

Pop N Fresh
09-21-2003, 09:27 AM
I'd actually be interested in knowing this to. Anyone have the secret?

NitroGL
09-21-2003, 09:46 AM
It would be breaking the NDA if someone did say.

The GLslang is still incomplete anyway.

Myagi
09-21-2003, 11:52 AM
Well I didn't sign any NDA. However I don't remember what the EULA says about reverse engineering (if you can call looking at a file in a hex editor that) so I'll refrain from posting anything right now except that my problem is solved for the moment.

Tom78
09-24-2003, 01:17 PM
Breaking the NDA, wow !
It's only an extension and it is one that people are waiting for a very long time. Why can't ATI say : use it, it's not ready, but for testing it's good enough. I have rewritten my shadercode twice now and i will have to do it a third time for GLSlang. I would like to start the conversion NOW and not in 3 months. Of course, i could buy a graphics board from 3DLabs and use their drivers for GLSlang, but i am only a student and i don't have the money for such actions. Why is it possible for 3D Labs to push GL2 and GLSlang and ATI or NV can't ? This holy graphics war begins to suck very much.
Please ATI, give also some students the ability to access this functions without searching around in the dlls. Where is the problem ? We are wasting time by searching around in the dlls.

cu
Tom

[This message has been edited by Tom78 (edited 09-24-2003).]

Jan
09-24-2003, 03:34 PM
If you worked at ATI or nVidia, would you want others to use (and judge!) a feature, which might be far from perfect or even working?

I mean, do you show other people your cool new lighting effect demo (or whatever) when there is currently a bug in the code and it doesnīt work properly? I wouldnīt. I would wait until it is working very well, because then other people wonīt say "Hey, thatīs crap".

And additionally, they certainly donīt want their support-center to be flooded with e-mails, which describe problems they are very aware of, because it has not been finished yet.

On the other hand, i can understand you, of course. I would be happy, if i could start coding with 2.0 right now.

Jan.

DopeFish
09-24-2003, 07:32 PM
More than likely they havent finished coding it into the drivers. If you had drivers supporting it, what would you do if half the code didnt work anyway because its incomplete?

Either wait until they have finished coding it, or get into their beta program.

Tom78
09-24-2003, 11:19 PM
Originally posted by DopeFish:
Either wait until they have finished coding it, or get into their beta program.

That's interesting. Can individual programers and students join the beta program ?

cu
Tom